Memory Encoding & Retrieval – From Levels of Processing to Practical Mnemonics

Level of Processing Theory

Craik & Tulving (1975) proposed that the durability of a memory trace depends on the depth at which incoming information is processed.

• Shallow (visual-orthographic) ➜ minimal analysis of letter shape.
• Intermediate (phonological) ➜ analysis of sound/rhyme.
• Deep (semantic) ➜ analysis of meaning and contextual fit.

Underlying claim: $\text{Deeper Processing} \; \Rightarrow \; \text{Better Encoding} \; \Rightarrow \; \text{Higher Recall}$

Classic Demonstration: Three Encoding Tasks

1. Capital-letter check
   “TOMORROW” – All caps? (Answer: No)
2. Rhyme check
   “GONE” – Rhymes with “done”? (Answer: Yes)
3. Sentence fit (semantic)
   “I put the __ in the taxi’s trunk.” – Does “carpet” fit? (Sometimes Yes)

After an unexpected recall test, results typically show:
$\text{Percent Recalled}<em>{\text{visual}} \ll \text{Percent Recalled}</em>{\text{rhyme}} \ll \text{Percent Recalled}_{\text{semantic}}$

Critiques & the Idea of Transfer-Appropriate Processing (TAP)

• "Depth" is circularly defined: a process is deep if it yields good memory – but we know it is good because it was remembered.
• Memory performance also depends on compatibility between encoding operations and the demands of the retrieval test.

Experimental twist:
– Encode word list with either meaning task or rhyming task.
– Later test with standard recognition (semantic) or *rhyme-cue test.*

Findings:
• Recognition test (semantic) favours prior semantic processing.
• Rhyme test reverses pattern – rhyming at encoding now wins.

Principle: $\text{Memory Best} \; \text{when} \; \boxed{\text{Encode Processes} = \text{Retrieve Processes}}$
This is Transfer-Appropriate Processing; depth per se is not sufficient.

Additional Factors That Enhance Encoding

Creating Rich Connections

• Link new material to existing knowledge, experiences, or other memories.
• Example cue for “chicken”: “The giant chicken jumps into the water and devours a red frog.”
– Vivid, bizarre, multi-sensory ➜ stronger retrieval route.

Imagery

• Paired associate list “boat – tree.”
– Simple repetition → ≈ $5\%$ recall.
– Form a striking image (a boat stranded atop a tree after a flood) → ≈ $15\%$ recall (≈ 3× improvement).

Self-Reference Effect

• Task 1: judge word length → ≈ $15\%$ recall.
• Task 2: “Does this word describe you?”
– Self-referential judgment boosts recall dramatically (often > $35\% – 40\%$).
Reason: the self forms an elaborated, well-organised knowledge network.

Generation Effect

• Study pairs you generate ("king – cr____"→“crown”) vs. pairs you simply read (“king – crown”).
• Generating boosts recall by ~$28\%$ relative to rote exposure.

Organisation & Chunking

• People spontaneously cluster recall into categories (e.g., furniture, professions).
• Provide structure explicitly ("These 12 words are minerals: sapphire, emerald, …") → recall can triple.
• Organised slides with titles & “bottom-line” summaries exploit the same mechanism.

Meaningful Frameworks

• Bransford & Johnson style passage: text about laundry recalled poorly when read without context; two-word cue “doing laundry” before reading restructures comprehension and massively improves later memory.

Testing Effect (Retrieval Practice)

Roediger & Karpicke (2006):

• Condition A – Reread passage.
• Condition B – Read once, then free-recall test.

Retention Results
– $\Delta<em>{5\;\text{min}} \approx 0$ (no advantage). – $\Delta</em>{2\;\text{days}}$ & $\Delta_{1\;\text{week}}$: tested group far outperforms reread group.
Interpretation: active retrieval consolidates memory traces.

Retrieval: Finding Stored Memories

Retrieval Failure vs. Decay

Evidence suggests traces rarely vanish; more often they become inaccessible.
Phenomenon: Tip-of-the-Tongue (TOT) – high feeling-of-knowing without access.

Retrieval Cues

• Imagery cues (giant chicken) – vivid & bizarre → easier search.
• Spatial cues – remembering where you studied a fact can trigger recall.
• Explicit category cues: give learners the categories (“birds, furniture, professions”) → recall jumps from ≈ $40\%$ to much higher.

Method of Loci (Memory Palace)

Step-by-step route through a familiar environment (home → hallway → elevator → street …).
– Place each to-be-remembered item at a specific landmark: hallway full of milk, elevator showering pasta, sidewalk covered with walnuts, tree sprouting bananas, etc.
– Later, mentally walk the route; each locus supplies the cue.
Ideal for speeches (ordering main topics) and shopping lists.

Practical Study Recommendations Drawn from the Lecture

1. Engage deeply & compatibly – predict exam question format and match your encoding to it.
2. Build connections – link concepts to prior courses, real-world examples, or personal anecdotes.
3. Use vivid imagery; the odder the better.
4. Ask “How does this relate to me?” – harness the self-reference effect.
5. Teach or explain material to peers; generating & organising thoughts reveals gaps.
6. Organise notes with headings, concept maps, categories, and summaries.
7. Supply yourself with a framework before diving into dense passages.
8. Replace some rereading with retrieval practice – flashcards, practice tests, peer quizzing.
9. Prepare retrieval cues (mnemonics, loci, acronyms) ahead of time.
10. Remember: failure to recall usually reflects cue mismatch, not destroyed memories – rebuild access paths.